Abstract Background Astrocytes regulate neuronal function, synaptic formation and maintenance partly through secreted extracellular vesicles (EVs). In amyotrophic lateral sclerosis (ALS) astrocytes display a toxic phenotype that contributes to motor neuron (MN) degeneration. Methods We used human induced astrocytes (iAstrocytes) from 3 ALS patients carrying C9orf72 mutations and 3 non-affected donors to investigate the role of astrocyte-derived EVs (ADEVs) in ALS astrocyte toxicity. ADEVs were isolated from iAstrocyte conditioned medium via ultracentrifugation and resuspended in fresh astrocyte medium before testing ADEV impact on HB9-GFP^+ mouse motor neurons (Hb9-GFP^+ MN). We used post-mortem brain and spinal cord tissue from 3 sporadic ALS and 3 non-ALS cases for PCR analysis. Findings We report that EV formation and miRNA cargo are dysregulated in C9ORF72-ALS iAstrocytes and this affects neurite network maintenance and MN survival in vitro. In particular, we have identified downregulation of miR-494-3p, a negative regulator of semaphorin 3A (SEMA3A) and other targets involved in axonal maintenance. We show here that by restoring miR-494-3p levels through expression of an engineered miRNA mimic we can downregulate Sema3A levels in MNs and increases MN survival in vitro. Consistently, we also report lower levels of mir-494-3p in cortico-spinal tract tissue isolated from sporadic ALS donors, thus supporting the pathological importance of this pathway in MNs and its therapeutic potential. Interpretation ALS ADEVs and their miRNA cargo are involved in MN death in ALS and we have identified miR-494-3p as a potential therapeutic target. Funding: Thierry Latran Fondation and Academy of Medical Sciences. Keywords: Astrocytes, Neurodegeneration, Gene therapy, Axonal growth, Extracelular vesicles, miRNA, Amyotrophic lateral sclerosis __________________________________________________________________ Research in context. Evidence before this study Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease characterized by motor neuron death, however, astrocytes contribute to the neurodegenerative process. Recent studies have demonstrated that astrocytes from ALS patients can induce motor neuron death through secreted factors. Several studies have focused on secreted proteins without identifying a culprit, however. Recent studies have highlighted the importance of microRNAs in cell-to-cell communication and have implicated miRNAs in the cross-talk between motor neurons and astrocytes in ALS. Added value of this study In this study we tested the effect of extracellular vesicles (EVs) secreted by astrocytes from ALS patients as opposed to non-ALS donors on motor neuron survival. We found that EVs secreted by ALS astrocytes induce motor neuron death nearly at the same levels as the conditioned medium. Moreover, we found that the main component of EVs are miRNAs, molecules that negatively regulate gene expression. Through a microarray study we identified several dysregulated miRNA that are likely to disrupt neuronal function once taken up by motor neurons. In particular, we focused on miR-494-3p that downregulates various genes including semaphorin3A, which is involved in axonal growth and maintenance. Implications of all the available evidence This study has uncovered the functional importance of EV secretion dysregulation that had been previously described in C9ORF72-ALS samples. Through the use of patient-derived astrocytes we have identified a number of new potential therapeutic targets for ALS that can be manipulated to restore neuronal function and prevent motor neuron death. Although challenging to use in gene therapy due to their ability to target several transcripts at once, our study provides evidence that manipulation of individual miRNAs can lead to significant beneficial downstream effects in vitro to be validated in vivo. Alt-text: Unlabelled Box 1. Introduction Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by motor neuron (MN) degeneration. The mechanisms and sequence of events leading to MN death are still widely unknown, but the observation that the first pathophysiological changes observed in patients involve neuromuscular junction (NMJ) disruption have given rise to the theory known as the ‘dying-back’ hypothesis [[45]1]. Dysregulated RNA metabolism is strongly implicated in this pathogenesis [[46]2], as demonstrated by the many ALS-linked genes encoding proteins involved in RNA metabolism, such as TARDBP and FUS [[47]3]. In particular, microRNA (miRNA) metabolism dysregulation has also been implicated in ALS [[48][4], [49][5], [50][6]]. In addition, polymorphic (G4C2)n hexanucleotide repeat expansions within the C9ORF72 gene are the most common genetic cause of ALS and frontotemporal dementia [[51]7,[52]8]. They are known to mediate neurotoxicity through multiple mechanisms including alterations of pre-mRNA processing [[53]9,[54]10], along with dysregulations of autophagy, protein homeostasis and vesicle trafficking [[55][11], [56][12], [57][13]]. Although ALS is characterized by MN degeneration, rodent studies have demonstrated that astrocytes dictate disease progression in vivo [[58]14]and patient-derived astrocytes are toxic towards wild-type MNs through both cell-to-cell contact and secreted factors in vitro [[59][15], [60][16], [61][17]]. While several hypotheses have been put forward [[62]18,[63]19], there is no consensus on the nature of these toxic factors. Under normal physiological conditions, astrocytes regulate many neuronal functions including axon maintenance [[64]20], and at least part of this communication is regulated through secreted extracellular vesicles (EVs) [[65]21,[66]22]. Specifically, EV miRNA cargo can modulate neuronal and astrocytic function in health and disease [[67]23,[68]24]. In the context of ALS, recent rodent-based studies have implicated cell-secreted miRNA signaling in a number of pathogenic processes, from excitotoxicity [[69]24] to neuromuscular junction disruption [[70]25]. Based on our previous research, we sought to determine whether astrocyte-derived extracellular vesicles (ADEV) from ALS patients contain distinct and/or altered levels of miRNAs that would account for the astrocyte toxicity reported against MNs [[71]16]. Here we show that astrocytes derived from C9ORF72-ALS patients have impaired EV formation. We also show that the miRNA cargo of these EVs is specific to astrocytes compared to the fibroblasts of origin and significantly differs from healthy control astrocytes. Specifically, we have identified miR-494-3p as a key regulator of Semaphorin 3A (Sema3A) and other molecules involved in axonal maintenance, with dramatic consequences on axonal/neurite length and motor neuron survival in vitro. Mir-494-3p dysregulation was also detected in the cortico-spinal tract isolated from post-mortem ALS biopsies, thus corroborating the importance of this pathway in disease and supporting its validity for future therapeutic development. 2. Materials and methods 2.1. Human sample ethics statement All skin biopsy donors ([72]Table 1) provided informed consent before sample collection (University of Sheffield, Study number [73]STH16573, Research Committee reference 12/YH/0330 or Coriell Institute). Table 1. Details of fibroblast donors. Cell line Diagnosis Mutation Age at collection (y) Gender Source Identifier AG08620 Non-ALS – 64 Female Coriell Institute RRID:CVCL_4L17 155 Non-ALS – 42 Male UoS RRID:CVCL_UF81 3050 Non-ALS – 55 Male UoS RRID: CVCL_UH66 78 fALS C9ORF72 68 Male UoS RRID:CVCL_UF84 183 fALS C9ORF72 51 Male UoS RRID:CVCL_UF85 201 fALS C9ORF72 66 Female UoS RRID:CVCL_UF86 [74]Open in a new tab 2.2. Conversion of skin fibroblasts to induced neural progenitor cells (iNPCs) Skin fibroblasts from 3 controls and 3 C9-ALS patients ([75]Table 1) were reprogrammed as previously described [[76]16]. Briefly, 10^4 fibroblasts were grown in one well of a six-well plate. Day one post seeding the cells were transduced with retroviral vectors containing Oct 3/4, Sox 2, Klf 4 and c-Myc. Following one day of recovery in fibroblast medium, DMEM (Gibco, Waltham, MA, USA) and 10% FBS (Life Science Production, Bedford, UK) the cells were washed 1× with PBS and the culture medium was changed to NPC conversion medium comprised of DMEM/F12 (1:1) GlutaMax (Gibco, Waltham, MA, USA), 1% N2 (Gibco, Waltham, MA, USA), 1% B27 (Gibco, Waltham, MA, USA), 20 ng/ml FGF2 (Peprotech, Rocky Hill, NJ, USA), 20 ng/ml EGF (Peprotech, Rocky Hill, NJ, USA) and 5 ng/ml heparin (Sigma, St. Louis, MO, USA). As the cell morphology changes and cells develop a sphere-like form they can be expanded into individual wells of a six-well plate. Once an iNPC culture is established, the media is switched to NPC proliferation media consisting of DMEM/F12 (1:1) GlutaMax, 1% N2, 1% B27, and 40 ng/ml FGF2. 2.3. iAstrocyte differentiation and maintenance iAstrocytes were yielded as previously described [[77]9,[78]16]. Briefly, iNPCs were switched to astrocyte proliferation media, DMEM (Fisher Scientific, Hampton, NH, USA), 10% FBS (Life science production, Bedford, UK), 0.2% N2(Gibco, Waltham, MA, USA). Cells were grown in 10 cm dishes coated with fibronectin for 7 days unless otherwise stated. For Nanoparticle Tracking Analysis experiments, astrocyte medium was switched to EV free medium (DMEM (Gibco, Waltham, MA, USA) and 10% (v/v) knockout serum replacement (Gibco, Waltham, MA, USA)) 24 h before medium collection (day 6 of differentiation). Pre-conditioned media was collected from the iastrocytes at day 7 and centrifuged at 300 ×g for 10 min, prior to evaluation using the ZetaView (Particle Metrix, Meerbusch, Germany). 2.4. MN monocultures with EVs from iAstrocytes Murine Hb9-GFP^+ MN cultures were prepared from mouse embryonic stem cells (mESC) containing a GFP gene controlled by the MN-specific promoter Hb9 (kind gift from Thomas Jessel, Columbia University, New York). Hb9-GFP mESC were maintained by culturing on primary mouse embryonic fibroblasts (Merck, Burlington, MA, USA) in mESC media (KnockOut DMEM (Gibco, Waltham, MA, USA), 15% (v/v) embryonic stem-cell FBS (Gibco, Waltham, MA, USA), 2 mM l-glutamine (Gibco, Waltham, MA, USA), 1% (v/v) nonessential amino acids (Gibco, Waltham, MA, USA) and 0.00072% (v/v) 2-mercaptoethanol (Sigma, St. Louis, MO, USA)). mESCs were then differentiated into MN-enriched cultures via embryoid bodies (EBs). Briefly, mESCs were lifted using trypsin, resuspended in EB medium (DMEM/F12 (Gibco, Waltham, MA, USA), 10% (v/v) knockout serum replacement (Gibco, Waltham, MA, USA), 1% N2 (Gibco, Waltham, MA, USA), 1 mM l-glutamine (Gibco, Waltham, MA, USA), 0.5% (w/v) glucose (Sigma, St. Louis, MO, USA) and 0.0016% (v/v) 2-mercaptoethanol (Sigma, St. Louis, MO, USA)) and seeded into non-adherent Petri dishes. EB media was replenished every day, and 2 μM retinoic acid (Sigma, St. Louis, MO, USA) and 0.5 μM smoothened agonist (Sigma, St. Louis, MO, USA) were added daily from day 2 to day 7 post-seeding to induce mESC differentiation into MNs. After 7 days of differentiation, EBs were dissociated using 200 U/ml papain (Sigma, St. Louis, MO, USA). Hb9-GFP^+ MN were sorted using FACS Aria machine and 40,000 cells/well were plated in 96 well plates (Cellstar, Sigma, St. Louis, MO, USA) pre-coated with laminin 1:200 (Sigma, St. Louis, MO, USA) and polyornithine 1:1000 (Sigma, St. Louis, MO, USA) in PBS. MNs were cultured in 100 μl of MN media (Knockout DMEM, F12 medium, 10% Knockout Serum Replacement, 1 mM l-glutamine, 0.5% (w/v) glucose, 1% N2, 0.0016% (v/v) 2-mercaptoethanol, 20 ng/ml BDNF (Peprotech, Rocky Hill, NJ, USA), 40 ng/ml CNTF (Peprotech, Rocky Hill, NJ, USA) and 20 ng/ml GDNF (Peprotech, Rocky Hill, NJ, USA)) for 24 h before treatment with either complete astrocyte conditioned media or isolated extracellular vesicles (EVs) appropriately diluted in fresh iastrocyte medium (DMEM (Gibco, Waltham, MA, USA), 10% FBS (Life Science Production, Bedford, UK) and 0.2% N2(Gibco, Waltham, MA, USA)) in order to keep the same concentration of EVs present in the conditioned medium. Each treatment well, comprised one part MN media and two parts either complete astrocyte conditioned media or isolated EVs or isolated EVs diluted in complete MN medium (including growth factors BDNF/CNTF/GDNF at the concentrations specified above). MNs were imaged using an INCELL analyser 2000 (GE Healthcare, Chicago, IL, USA) 24 h after seeding (to confirm the number of cells before treatment), and then every 24 h onwards for 3 days. The number of viable MN (defined as GFP+ motor neuronal cell bodies with at least 1 axon) were analysed using the Columbus™ Data Storage and Analysis System (RRID:[79]SCR_007149; Perkin Elmer, Waltham, MA, USA). The percentage MN survival was calculated as the number of viable motor neurons at day 3 as a percentage of the number of viable MN at day 0 pre-treatment. As for treatment with miRNA mimics (MirVana, ThermoFisher, Waltham, MA, USA), scramble and hsa-miR-494-3p, were added to the conditioned medium or ADEVs on the day of treatment. 2.5. EV preparation EVs were isolated from conditioned medium by ultracentrifugation at 100,000 ×g for 90 min at 4 °C using a 70ti rotor and Beckman Coulter Ultracentrifuge after initial collection and centrifugation for 10 min, 300 ×g at room temperature and filtration through a 0.22 μm filter to remove cell debris. The supernatant was then removed and the EV pellet resuspended in 300 μl DEPC treated PBS. 2.6. NTA Nanoparticle tracking analysis was conducted using the ZetaView (RRID:[80]SCR_016647; Particle Metrix, Meerbusch, Germany) and its respective software (RRID:[81]SCR_016647; ZetaView 8.03.08.03). Prior to use the instrument was calibrated using polystyrene beads (100 nm). Conditioned media samples taken from human induced astrocyte cultures were loaded into the ZetaView cell. Nanoparticle tracking analysis measurements were recorded at 11 different positions and three cycles of readings were documented for each position. Following the withdrawal of any outlier positions the ZetaView software calculated the mean, median and mode sizes in addition to the concentration of particles within the sample. 2.7. EV transmitted electron microscopy For immuno-gold EM, a 5 μl drop of resuspended EVs was deposited onto the grid and adsorbed for 20 min. Grids were washed in 100 μl drops of PBS 3 times, followed by blocking for 10 min in 100 μl of blocking buffer consisting of 5% horse serum in PBS. A 5 μl drop of Ms. CD63(TS63, Invitrogen; Cat# 10628D, RRID:[82]AB_2532983; Carlsbad, CA, USA) of 20 μg/ml was incubated on the grid for 20 min, washing the grid in PBS 3 times at the end of incubation. 5 μl of anti-mouse IgG-10 nm gold (Cat#ab39619; RRID:[83]AB_954440; Abcam, Cambridge, UK) in 5% horse serum was adsorbed onto the grid for 20 min, and washed with PBS. The immunoreaction was post-fixed with EM grade 3% glutaraldehyde/formaldehyde for 5 min, and the sample was contrasted with 2% uranyl acetate for 60 s and washed 3 times with distilled water before drying overnight. Samples were imaged with a Fei Tecnai 2000 electron microscope at 80 kV. 2.8. RNA isolation and quantitative RT-PCR RNA was harvested using the RNeasy Plus mini Qiagen kit (Qiagen, Germantown, MD, USA) and total RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA) in accordance with the manufacturer's instructions. Real-time quantitative PCR reactions were conducted using 2× SYBR Green qPCR Master Mix (Low ROX) (Bimake, Houston, TX, USA) and assays were run on a Stratagene Mx3000P™ Real Time Thermal Cycler (Agilent Technologies Ltd., Santa Clara, CA, USA). Mouse Sema3A RNA levels were detected and relative quantification calculated using the 2−∆∆CT method. The following primers were utilised: mouse Sema3A Fw 5’-TGTACTCTGGAACTGCTGCG-3′; Rv 5’-TCTCTGGGATGAGATGGGCA-3′; mouse GAPDH Fw 5’-GCTACACTGAGGACCAGGTTGTCT-3′; Rv 5’-AGCCCCGGCATCGAA-3′; human Sema3A Fw 5’–CACCATCACCCCATCAGGAC-3′; Rv 5’-CTCTGGGATGAGATGGGCAC-3′ and human RPL13A Fw 5’-CAAGCGGATGAACACCAACC-3′; Rv 5’-TTTTGTGGGGCAGCATACCT-3′. For TaqMan qPCR 6 genes associated with EV formation were selected ([84]Table 2) and PCR was performed using PrimeTime qPCR 5′ nuclease assays (IDT technologies, Coralville, IA, USA) as described above. Gene expression values were determined using the ddCt calculation following normalization to β-Actin expression, also evaluated using the associated PrimeTime qPCR 5′ nuclease assay. Table 2. Primers and probes for TaqMan qPCR. Primer name Primer sequence (5′ – >3′) Probe sequence (5′ – >3′) ALIX-F GAAGCACAGGTGGTGGAG 56FAM/CATGGTTCT/ZEN/TGGCGCTGGAGTTG ALIX-R CAGCAGGAGGACATGCAC TSG101-F TTTTCCAGAGCAGAACTGAGT 56FAM/AACCTCGGC/ZEN/TACTTCTTGATCTAAACGG TSG101-R GAAAAAGGGTCACCAGAAACTG CHMP2B-F TCGTCATCAGAACCGTCAAAG 56FAM/CAGAAGGAA/ZEN/AACATGAAAATGGAAATGACTGAAGA CHMP2B-R AGGCAGTTAACAAGAAGATGGA CHMP4B-F GGAACATTTGGTAGAGGGACTG 56FAM/TGGCGGAAT/ZEN/TAGAAGAACTAGAACAGGAGG CHMP4B-R AGGGTTTGGAGAAGAGTTTGAC VPS4A-F CTCTTGCCCAAAGTCCTCTG 56FAM/TTCTTCACT/ZEN/TTCAGGAGGTCGTCTGC VPS4A-R TCTTAGAGCCTGTGGTTTGC β-Actin-F CCAGTGGTACGGCCAGA 56FAM/CCATGTACG/ZEN/TTGCTATCCAGGCTGT β-Actin-R GCGAGAAGATGACCCAGAT [85]Open in a new tab For miRNA analysis, total RNA was isolated using TRIzol™ Reagent (Invitrogen, Carlsbad, CA, USA) as per the manufacturer's guidelines TaqMan Small RNA Assay (20×) for hsa-miR-103 and hsa-miR-494-3p and TaqMan Universal PCR Master Mix II (2×) (both Life Technologies Ltd., Carlsbad, CA, USA) were used for real-time quantitative PCR reactions and the assays were conducted on a Stratagene Mx3000P™ Real Time Thermal Cycler (Agilent Technologies Ltd., Santa Clara, CA, USA). 2.9. Microarray analysis Total RNA was extracted from EV pellets derived from the conditioned media of C9ORF72 patient and control patient induced astrocytes using the Direct-Zol RNA mini-prep (Zymo Research, Irvine, CA, USA) as per the manufacturer's instructions. RNA was labelled using the FlashTag biotin HSR RNA labelling kit (Affymetrix, Santa Clara, CA, USA) according to the manufacturer's instructions. Briefly, 2 μl of RNA Spike Control Oligos were added to 8 μl of RNA, whilst ATP mix was diluted 1:500 in 1 mM Tris. 5 μl of Master Mix was added to 10 μl RNA/Spike Control Oligos and incubated at 37 °C heat for 15 min (Poly (A)-tailing reaction). To the Poly (A)-tailed RNA 4 μl of 5× FlashTag Biotin HSR Ligation Mix and 2 μl of T4 DNA Ligase was added and following a 30 min incubation at room temperature the reaction was stopped by adding 2.5 μl of HSR Stop Solution. Labelled RNA from each sample was then prepared for hybridisation. The hybridisation cocktail, comprised of 110.5 μl of Hybridization Master Mix and 21.5 μl of biotin-labelled sample was incubated at 99 °C for 5 min and then 45 °C for 5 min. The sample mixture was then loaded into an array and placed into a hybridization oven and incubated at 48 °C and 60 rpm for 16 to 18 h. Arrays were then washed and stained using Stain Cocktails 1/2 and scanned with the Affymetrix GeneChip Scanner 3000. Affymetrix Expression Console was used to evaluate the hybridisation efficacy and probe set expression of the GeneChip miRNA Array 4.0. Differential miRNA expression between samples were analysed using Affymetrix Transcriptome Analysis Console for RNA species annotated within the human genome. The Affymetrix Expression Console was then used to evaluate probe set expression after normalization using the detected above background (DABG) method. DABG is a detection metric where the intensity from each perfect match probe is compared to a distribution of background probes and provides a probability that value of the signal intensity if part of the background noise [[86]26]. Different normalization methods and cutoffs are used in miRNA expression analysis depending on the expression level of the miRNAs in the sample analysed [[87]27,[88]28]. After normalization and after applying a pvalue cutoff ≤0.05, differentially expressed miRNAs were evaluated after applying fold changes ≥2, 1.5 and 1.2. Due to the low number of miRNA displaying a fold change ≥2, fold-change of 1.5 and 1.2 were used as an arbitrary threshold to perform pathway enrichment analysis in order to identify the pathways that are overall affected by ALS ADEVs. To elucidate pathways targeted by dysregulated miRNAs three online analytical tools were used, Database for Annotation, Visualization and Integrated Discovery (DAVID) [[89]29], DIANA-mirPath [[90]30] and miRSystem [[91]31]. MiRSystem generates a list of enriched pathways that are regulated by queried miRNAs by incorporating miRNA expression values and matching miRNAs with the latest annotation [[92]31]. It combines seven different algorithms and two validated databases to identify target genes and uses five pathway databases to characterize the enriched pathways. DIANA-miRPath performs miRNA pathway analysis and allows queried miRNAs and target genes to be visualized on pathway maps [[93]30]. DAVID is a widely cited and used tool that discovers enriched functional-related gene groups and allows them to be visualized on pathway maps [[94]29]. DAVID and DIANA-miRPath utilize enrichment p-values while miRSystem uses ranking scores based on the affinity of the interactions between miRNAs and target genes [[95]31]. Five pathways that were common to more than two tools were found. From the five common pathways listed for up and downregulated miRNAs, two pathways that were of our interest were selected and genes that are involved in those pathways were found. The most enriched miRNAs were found by selecting genes, involved in the pathways of interest that were commonly found between the analytic tools. 2.10. Post mortem tissue Autopsy donations to the Sheffield Brain Tissue Bank were performed with the written consent of the next of kin for the use of tissues for scientific research. Slices of brain and segments of spinal cord were frozen on copper plates in liquid nitrogen vapour and stored at −80 °C. Six samples were retrieved from deep freezing ([96]Table 3), and samples of motor cortex, spinal cord and lateral corticospinal tract taken by a qualified neuropathologist. Table 3. Details of post-mortem tissue donors, including gender, age, post-mortem delay (PMD) and diagnosis. Case ID Gender Age PMD Diagnosis C1 F 63 31 h Carcinoid Tumor C2 M 63 25 h Mesolthelioma C3 M 67 63 h Hepatocellular carcinoma sALS1 M 66 40 h sALS sALS2 F 63 36 h sALS sALS3 M 66 19 h sALS [97]Open in a new tab The remaining nervous system tissue was formalin fixed for diagnostic confirmation and characterisation. The Sheffield Brain Tissue Bank Management Board gave ethical approval for use of tissue in this study under the provision to act as a Research Tissue Bank as approved by the Scotland Research Ethics Committee (ref. 08/MRE00/103). 2.11. Statistical analysis All statistical tests were conducted using Graph Pad Prism 7 software (RRID:[98]SCR_002798). Statistical analysis was performed by either Student's t-test, one-way ANOVA with Tukey post-hoc analysis, or two-way ANOVA with Sidak post-hoc analysis, depending on the number of variables in the respective experiment. All experiments were performed a minimum of three times. p < .05 was considered statistically significant. All p values and n values are documented in the figure legends. 3. Results 3.1. Conditioned medium and extracellular vesicles from C9orf72 astrocytes are toxic to MNs We previously showed that C9orf72 iAstrocytes derived through direct conversion of fibroblasts into iNPCs induced a decrease in survival in Hb9-GFP^+ mouse MNs in co-culture compared to non-ALS astrocytes [[99]16]. Here we tested the effect of iAstrocyte conditioned medium on MN monocultures, to determine if secreted factors are main players in C9-ALS astrocytic-mediated neurotoxicity as previously reported for other genetic subtypes [[100]15,[101]17]. Consistent with previous data, C9-ALS astrocyte conditioned medium treatment of MN monocultures resulted in 50–75% increased MN death compared to controls after 72 h ([102]Fig. 1a, b, e; p < .0001 (Two-way ANOVA)). Since extracellular vesicles (EVs) have been implicated in cell-to-cell communication in a number of neurodegenerative diseases [[103]32], we decided to determine their role in astrocyte-mediated MN death in C9-ALS. We isolated EVs from C9 iAstrocyte conditioned medium (10 ml) via ultracentrifugation, resuspended them in fresh non-conditioned astrocyte medium (10 ml) or complete MN medium with growth factors (10 ml) and we then treated the MN monocultures. Our data reveal that C9-EVs are equally toxic as the astrocyte conditioned medium ([104]Fig. 1c-e, p < .0001 (Two-way ANOVA)). Dilution of the EVs in complete MN medium containing BDNF, GDNF and CNTF still resulted in MN death, proving that EVs are toxic to MNs (Supplementary Fig. 1). We also report here that ADEVs isolated from control astrocytes are consistently, even if not statistically significant, less supportive compared to whole conditioned medium ([105]Fig. 1e) while C9-EVs are consistently less toxic than whole conditioned medium. This suggests that other factors not packaged into ADEVs contribute to MN health/death. Consistent with these findings, MN medium alone is not as supportive as astrocyte-conditioned medium from non-ALS controls ([106]Fig. 1e and Supplementary Fig. 2a, b), proving that astrocytes have an active role in supporting neurite growth and synaptic formation. Fig. 1. [107]Fig. 1 [108]Open in a new tab Conditioned medium (CM) and extracellular vesicles (EVs) from C9-ALS iAstrocytes induce MN death. Representative images of Hb9-GFP^+MN monocultures treated for 72 h with astrocyte conditioned medium from control (CTR) or C9-ALS (C9) astrocytes (a and b respectively) or EVs isolated from the same conditioned media (c and d respectively). Quantification of HB9-GFP^+MNs with axons performed after 3 days of treatment reveals that conditioned medium (solid bars) and EV treatment (checked bars) have similar effects on MN monocultures (e), with C9-ALS astrocyte medium and EVs inducing significantly lower MN survival. Survival of MN monocultures in MN medium (black solid bar) used as comparative reference (n = 3). Two-way ANOVA (p < .0001), 3xControls versus 3xC9-ALS, N = 3 per condition, error bar = SD. Scale bar = 20 μm. 3.2. EV biogenesis is impaired in C9orf72 astrocytes Having assessed the detrimental effect of EVs secreted by C9-ALS astrocytes, we sought to characterize their abundance in the conditioned medium. We used the ZetaView Nanoparticle Tracking Analysis (NTA) system to identify particle size and number in conditioned medium from 3 controls and 3 C9-ALS patients. Quantification revealed no difference in the overall size range of EVs secreted by controls or patients, typically ranging between 70 and 200 nm. Most particles range between 50 and 120 nm, with a clear peak at 100 nm, indicating that exosomes are the main component of the total EV pool ([109]Fig. 2a). We confirmed ADEV isolation by transmitted electron microscopy (TEM) and immunogold staining for CD63 ([110]Fig. 2b and c). Fig. 2. [111]Fig. 2 [112]Open in a new tab EV biogenesis is impaired in C9-ALS iAstrocytes. Representative size distribution graph (a) of ADEVs and TEM (b, c-i, c-ii) show that most ADEVs range between 50 and 120 nm. ADEVs are positive for CD63 (arrow heads) with immunogold labeling (c-iii). ADEV direct quantification using the ZetaView Nanoparticle Tracking System (d) shows a significant decrease in the number of particles secreted by C9 iAstrocytes (One-way ANOVA; p < .01). This suggests impairment in vesicle formation, as supported by TaqMan qRT-PCR data showing downregulation in a number of transcripts involved in EV biogenesis in C9 iAstrocytes (e). t-test (**p < .01; *** p < .001), 3xControls versus 3xC9-ALS, N = 3 per condition, error bar = SD. Interestingly, ADEV quantification showed that iAstrocytes from C9 patients secrete fewer vesicles than healthy individuals across a range of vesicle sizes ([113]Fig. 2d and [114]Fig. 3). It was recently reported that C9orf72 not only is involved in autophagy [[115]13], but also in vesicle trafficking [[116]11]. For this reason, we decided to assess the mRNA expression level of proteins involved in EV formation and processing. Consistent with the observed reduction in EV secretion, qRTPCR analysis showed that transcripts encoding for C9orf72, TSG101, CHMP4B and VSP4A are significantly downregulated in C9-ALS iAstrocytes ([117]Fig. 2e). Fig. 3. [118]Fig. 3 [119]Open in a new tab MicroRNAs regulating axonal maintenance are selectively dysregulated in C9 ADEVs. MiRNAs secreted by C9 and Control (CTR) iAstrocytes identify two separate groups in the hierarchical clustering analysis (a). The PCA plot (b) shows how EV-secreted miRNAs differentiate iAstrocytes (iA) and fibroblasts (Fib) on one axis and controls (CTR) and C9 patients (C9) on another axis. iA also show more dramatic differences between patients and controls compared to fibroblasts. Pathway analysis of the dysregulated miRNAs identifies axonal growth and maintenance as the most affected pathway (c). Downregulated (in blue) miRNAs target Semaphorins, RhoA and Rock, thus predicting an increase in these proteins, which would lead to growth cone collapse. Upregulated miRNAs (in red) target Ephs and Wwp1, which would lead to their downregulation. Wwp1 inactivates NogoA, thus this would also lead to axonal collapse. TaqMan qRT-PCR confirms significant downregulation of miR-494-3p, which negatively regulates Sema3A (d). N = 3 per group; error bar = SD. Two-tail unpaired t-test (p < .05). (For interpretation of the references to colour in this figure legend, the reader is